【作者】 胡赤兵； 楼军伟； 王季； 李贵子；

【Author】 HU Chibing~1,LOU Junwei~1,WANG Ji~2,LI Guizi~2 (1.School of Mechanical and Electronical Engineering,Lanzhou University of Technology, Lanzhou,730050 China;2.Key Laboratory of Mechanical Product Testing and Technical Evaluation of Gansu Province,Lanzhou,730030 China)

【机构】 兰州理工大学机电工程学院； 甘肃省机械产品检测与技术评价重点实验室；

【摘要】 针对常用的傅里叶变换、小波变换等在分析滚动轴承故障信号时存在难以度量复杂性的局限,采用经验模态分解(EMD)结合样本熵来改进。通过对不同损伤程度的滚动轴承信号应用样本熵和EMD样本熵的实际效果进行比较。发现样本熵值过于接近不易区分,而EMD样本熵值差别准确性更高,并且其变化趋势与故障信号随损伤变化的趋势一致。更多还原

【Abstract】 Aiming at commonly used Fourier Transform,Wavelet Transform etc were difficult to measure the complexity in analysising the rolling bearing signal,using EMD combined with sample entropy to improve.By comparing the actual effect of sample entropy and EMD sample entropy were used in different damage levels of rolling bearing singal.Finding that the sample entropy too close to distinguish,but the EMD sample entropy obviously different and higher exact and the changing trend was consistent with the damage changing trend.更多还原